The age of supergene manganese deposits in Katanga

and its implications for the Neogene evolution of the

African Great Lakes Region

Thierry De Putter, Gilles Ruffet, Johan Yans, Florias Mees

To cite this version:

Thierry De Putter, Gilles Ruffet, Johan Yans, Florias Mees. The age of supergenemanganese deposits in Katanga and its implications for the Neogene evolution of theAfrican Great Lakes Region. Ore Geology Reviews, Elsevier, 2015, 71, pp.350-362..

HAL Id: insu-01168514

https://hal-insu.archives-ouvertes.fr/insu-01168514

Submitted on 26 Jun 2015

HAL is a multi-disciplinary open accessarchive for the deposit and dissemination of sci-entific research documents, whether they are pub-lished or not. The documents may come fromteaching and research institutions in France orabroad, or from public or private research centers.

L’archive ouverte pluridisciplinaire HAL, estdestinée au dépôt et à la diffusion de documentsscientifiques de niveau recherche, publiés ou non,émanant des établissements d’enseignement et derecherche français ou étrangers, des laboratoirespublics ou privés.

https://hal.archives-ouvertes.frhttps://hal-insu.archives-ouvertes.fr/insu-01168514

ACC

EPTE

D M

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

1

The age of supergene manganese deposits in Katanga and its implications for

the Neogene evolution of the African Great Lakes Region

Thierry De Putter1*

, Gilles Ruffet2, Johan Yans

3, Florias Mees

1

1 Royal Museum for Central Africa, Geodynamics and Mineral Resources, 13

Leuvensesteenweg, B-3080 Tervuren, Belgium

2 CNRS (CNRS/INSU) UMR 6118, Géosciences Rennes, F-35042 Rennes Cedex, France and

Université de Rennes I, Géosciences Rennes, F-35042 Rennes Cedex, France

3 Université de Namur, Département de Géologie, NaGRIDD, 61 rue de Bruxelles, B-5000

Namur, Belgium

* corresponding author: tel.: +3227695430, +32478213103 (mobile); fax: +3227695432. E-

mail address: thierry.de.putter@africamuseum.be

Abstract

Supergene manganese deposits commonly contain K-rich Mn oxides with tunnel structure,

such as cryptomelane, which are suitable for radiometric dating using the 39

Ar-40

Ar method.

In Africa, Mn deposits have been dated by this method for localities in western and southern

parts of the continent, whereas only some preliminary data are available for Central Africa.

Here we present new 39

Ar-40

Ar ages for Mn oxide samples of the Kisenge deposit, in

southwestern Katanga, Democratic Republic of the Congo. The samples represent supergene

Mn oxide deposits that formed at the expense of primary Paleoproterozoic rhodochrosite-

dominated carbonate ores. Main phases of Mn oxide formation are dated at c. 10.5 Ma,

3.6 Ma and 2.6 Ma for a core that crosses a mineralized interval. The latter shows a decrease

in age with increasing depth, recording downward penetration of a weathering front. Surface

ACC

EPTE

D M

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

2

samples of the Kisenge deposits also record a ≥ c.19.2 Ma phase, as well as c. 15.7 Ma,

14.2 Ma and 13.6 Ma phases. The obtained ages correspond to distinct periods of

paleosurface development and stability during the Mio-Pliocene in Katanga. Because

Katanga is a key area bordered to the North by the Congo Basin and to the East by the East

African Rift System, these ages also provide constraints for the geodynamic evolution of the

entire region. For the Mio-Pliocene, the Kisenge deposits record ages that are not

systematically found elsewhere in Africa, although the 10.5-11 Ma event corresponds to a

roughly simultaneous event in the Kalahari Manganese Field, South Africa. The rest of the

Katanga paleosurface record differs somewhat from records for other parts of Africa, for

which older, Eocene ages have been obtained. This difference is most probably related to the

specific regional geodynamic context: uplift of the East African Plateau, with associated

erosion, and the opening of the East African Rift System at c. 25 Ma are events whose

effects, in the study area, interfere with those of processes responsible for the development of

continent-wide paleosurfaces.

Keywords: 39

Ar-40

Ar geochronology, cryptomelane, Katanga, Neogene, supergene deposits

1. Introduction

Mineral-rich areas in Africa have since long drawn the attention of geologists looking for

mineable ore deposits. A typical example of a rich mining area is the Katanga Province, in

the south-eastern part of the Democratic Republic of Congo (DRC). From the early 20th

century onwards, industrial mining in Katanga has exploited copper, cobalt and uranium,

besides other metals such as lead, zinc, and accessory metals associated with the main Cu-Co

deposits (e.g. germanium). Regional geological studies have for decades focused on the

formation of the main primary deposits: Neoproterozoic sediment-hosted stratiform copper

ACC

EPTE

D M

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

3

(SHSC) and cobalt deposits in the Katanga Copperbelt, and the Early Phanerozoic vein-type

Pb-Zn-(Ge) deposit of Kipushi. However, mining operations also took considerable benefit of

the wealth of secondary, oxidized, copper, cobalt and manganese surface deposits in the

Katanga Province. Until recently, the formation processes and age of these supergene

deposits were poorly constrained. It is now known that these oxidized minerals formed quite

recently, in the Neogene (Decrée et al., 2010; De Putter et al., 2010), although older ages

have been obtained for other parts of Africa (Table 1).

Supergene deposits form in a geodynamic context that is rather specific. They develop at the

expense of primary protores or ores when these are brought to the surface and become

exposed, for a sufficient period of time, to the action of meteoric agents. The availability of

meteoric fluids is a first order constraint to alter the primary deposit, followed by

precipitation of secondary mineral phases, in the form of carbonates, oxides and other

compounds. The geodynamic context is thus obviously relevant to understand and

characterize how and when the secondary deposits formed. The latter present some

significant differences with their primary precursors: (i) paleotopographical context is a

major factor to explain their formation and preservation; (ii) fluids involved in their

formation are percolating meteoric waters; (iii) the host rock is often modified, becoming

either indurated (by silica or iron oxide cementation) or more friable (by weathering), which

in the second case results in increased susceptibility to erosion and exposure to oxidizing

fluids; and (iv) iron oxides often play a major role in re-concentrating elements of economic

value (other than manganese). Another difference with primary ore deposits is that supergene

ones are often considerably richer in economic metals, hence having considerable economic

interest.

Supergene ore development has been investigated for several major African mining areas (cf.

Dill et al., 2013), especially in western and southern Africa (Table 1). In contrast, they have

ACC

EPTE

D M

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

4

been considerably less studied for Central Africa, including Katanga. This paper aims at

bridging this gap, by presenting 39

Ar-40

Ar results for supergene ore deposits of the Kisenge

area in south-western Katanga, where a major secondary Mn oxide accumulation presents an

occurrence of suitable materials for absolute age determination.

2. Setting

2.1. Geological and geomorphological context

The Kisenge area hosts a world-class manganese deposit, occurring in outcrops along an

East-West trending 6 km-long series of small hills.

The geological setting of the study area is overall poorly known. The area belongs to the

southern part of the Congo Craton, bordering the southern edge of the present-day Congo

Basin (Fig. 1). Basement rocks are Archean or Paleoproterozoic (> 2.1 Ga) schists, gneiss,

migmatites and amphibolites (Ledent et al., 1962; Lepersonne, 1974; Cahen et al., 1984). The

primary Mn ore at Kisenge is a rhodochrosite- and spessartine-bearing deposit (ca 50% MnO)

interlayered with mudstones and shales, of probable Late Paleoproterozoic age (Schuiling &

Grosemans, 1956; Marchandise, 1958; Doyen, 1974), whose nature is the subject of an

ongoing investigation (De Putter and Mees, unpublished data). The best estimate for the age

of the Kisenge primary deposit is presently based on the regional geological context and on

radiogenic Rb-Sr ages obtained for muscovite in the pegmatite body that crosscuts the

deposit. The 1.8-1.9 Ga Rb-Sr age range represents a minimum age estimate for the deposit

(Doyen, 1974), which is moreover compatible with ages of similar Paleoproterozoic Mn

deposits (cf. Roy, 2006; Kuleshov, 2011). Comparable Mn deposits are known elsewhere in

Africa (Burkina Faso, Ghana, Gabon, South Africa), as well as in Brazil, and are all

Paleoproterozoic in age (1.9 to 2.2 Ga ; Schneiderhahn et al., 2006; Nyame et al., 2008;

ACC

EPTE

D M

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

5

Chisonga et al., 2012). Some of them, including the Kisenge deposit, are located along the

margins of the proto-Congo Craton, which included the Gabon and São Francisco blocks

(Fernandez-Alonso et al., 2012).

The Mn carbonate orebody is capped by a thick supergene Mn oxide deposit (Polinard,

1932). This deposit can be considered to be the Mn-dominant equivalent of (Fe-dominant)

laterites, which occur extensively in Katanga (Beugnies, 1954; Alexandre, 2002). Alexandre

(2002) proposed ages for their multi-stage formation, but this was done exclusively in

reference to assumedly equivalent deposits in western Africa (e.g. Tardy and Roquin, 1998).

In West Africa, only the Tambao deposit (Burkina Faso) has been extensively studied

through 39

Ar-40

Ar geochronology (e.g. Hénocque et al., 1998; Colin et al., 2005; Beauvais et

al., 2008) and now constitutes a reference system for a regional analysis of Cenozoic

paleosurfaces (Beauvais et Chardon, 2013).

Kisenge is situated on a peneplain at ~1,100 m a.s.l., rising to the east to 1,200 m around

Mutshatsha and >1,500 m a.s.l. around Kolwezi (Fig. 1), with relict hills culminating some

tens of meters above the peneplain level (Polinard, 1932; Doyen, 1974; Alexandre, 2002).

The hills marking Mn ore occurrences at Kisenge culminate at ~30 to 35 m above the mean

peneplain level (Fig. 2). No geomorphological survey has been performed for the Kisenge

area, but studies in more eastern parts of Katanga show the presence of various planation

surfaces (De Dapper, 1991).

Kisenge is located ~250 km West of the Katanga Copperbelt, which is part of the Lufilian

fold-and-thrust belt (Cailteux et al., 2005; Dewaele et al., 2006; Hitzman et al., 2012).This

world-class Cu-Co deposit hosts ~50 % of the world's known reserve of mineable cobalt

(3.4 Mt Co metal content; USGS, 2009) together with important Cu deposits. The Katanga

ACC

EPTE

D M

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

6

Copperbelt is also known for its thick derived supergene Cu-Co deposits assigned to the Mio-

Pliocene based on an earlier preliminary study (Decrée et al., 2010; De Putter et al., 2010).

With its extensive cover of supergene ore deposits, formed on a Paleoproterozoic Mn deposit

to the West and on a Neoproterozoic Cu-Co deposit (Copperbelt) to the East, the Katanga

province allows precise dating of the development of a major topographic high at the

southern margin of the Congo Basin. In a broader context, the geodynamics of the Katanga

area during the Cenozoic is probably related to: (i) an overall extensional tectonic regime

related to the Mesozoic break-up of Gondwana and opening of the EARS; (ii) long-term

erosion and planation of the Lufilian fold belt, and (iii) regional mantle and crust movements

associated with the formation of the western branch of the EARS, which started about 25 My

ago (Roberts et al., 2012; Kipata et al., 2013; Linol et al, 2015a).

2.2. Cenozoic regional geodynamic context

Current knowledge of the Cenozoic uplift history of the Katanga region and surrounding

areas, namely the Congo Basin, the Western Branch of the East Africa Rift System (EARS)

and the East African Plateau (EAP), can be summarized by reviewing the main documented

or inferred events for the Eocene to Pliocene epochs (Fig. 3). Quantitative estimates of

landmass/surface uplift changes have been recently published for the Congo Basin (Linol et

al., 2015a), but they are still lacking for the study area, south of the Basin.

In the Eocene, between 45 and 35 Ma, the EAP experienced major uplift, as a result of

buoyancy and melt generation beneath the region (Ebinger et al., 1993; Wichura et al., 2010).

During the same period, the Congo basin did not experience significant vertical movement

and possibly hosted a large lake at some stage (Peters and O‟Brien, 1999; Lavier et al., 2001;

Goudie, 2005; Guillocheau et al., 2015).

ACC

EPTE

D M

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

7

In the Oligocene, from around 25 Ma onward, both branches of the EARS opened (Pik et al.,

2008; Roberts et al., 2012). Around the same time, uplift of the Angolan High and the Congo

Basin resulted in a sharp increase in sedimentation rates offshore in the Congo fan (Lavier et

al., 2001; Anka et al., 2009 and 2010).

In the early to middle Miocene (c. 17-12 Ma; Wichura et al., 2010), the EAP experienced

another major uplift. To the West, the Congo Basin had a similar evolution somewhat later

(Lavier et al., 2001; Anka et al., 2009, 2010).

For the middle to late Miocene, a stable tectonic setting is recorded for the EAP, which

allowed the development of an abundant and stable tree cover in the c. 10 to 8 Ma interval

(Bonnefille, 2010). In the Congo Basin, uplift rates increased sharply after 11 Ma (Lavier et

al., 2001; Anka et al., 2009, 2010). Stable isotope studies on fossil eggshells suggest that the

entire Congo Basin area experienced an arid climate during the middle to upper Miocene

(Senut et al., 2009).

In the late Miocene (c. 7.5 Ma), major rifting took place in the Western branch of the EARS

(Pasteels et al., 1989), with associated culmination of volcanic activity. The uplift rate of the

Congo basin reached a maximum in the c. 11 to 5 Ma interval (Lavier et al., 2001;

Guillocheau et al, 2015).

In the Pliocene, the EAP still experienced uplift while to the west the uplift rate of the Congo

Basin sharply decreased (Lavier et al., 2001).

In the course of this Cenozoic history, and even since the end of the Pan-African orogeny (c.

450 Ma), the Eastern part of Katanga Plateau was apparently never covered by a thick

sediment cover. Mesozoic sediments in particular are lacking, in contrast to the thick

sequences deposited in the adjoining Congo Basin. Cenozoic deposits (Kalahari Series) occur

ACC

EPTE

D M

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

8

only sporadically in the Copperbelt area, but they become more continuous and thicker

toward the West, in the western DRC and Angola (Linol et al., 2015b).

During the Cenozoic, the Katanga Plateau occupied a stable position between two major

morphotectonic zones: the Congo Basin to the North and Northwest and the EARS to the

East and Northeast. This stability allowed the formation of an extensive lateritic cover on the

southern Katanga Plateau, as described by Alexandre (2002). The latter recognized four types

of laterites in the region, each attributed to a different period (Early Eocene, Eocene-

Oligocene, Miocene, Pliocene). However, no absolute age information was available to the

author in support of this proposed chronology.

3. Materials

3.1. Analysed mineral phase

The most suitable Mn oxide mineral for 39

Ar-40

Ar study is cryptomelane (K(Mn4+

7Mn3+

)O16),

which is the K-rich (>5%) end-member of a group of Mn oxides with tunnel structure

(coronadite group, hollandite supergroup) with A2+

[M4+

6M3+

2]O16 or A+[M

4+7M

3+]O16 as

general formula, with A = K, Na, Pb, Ba, Sr, and M3+

and M4+

= Mn (Biagioni et al., 2013).

These Mn oxides can have a composition that is intermediate between that of the end-

members of this group, resulting in variable K contents, also at the scale of aliquots used for

39Ar-

40Ar study. Moreover, other manganese oxide minerals may be present within analysed

grains, such as pyrolusite (MnO2) and lithiophorite ((Al,Li)(Mn4+

,Mn3+

)2O2(OH)2), as

confirmed by SEM-EDS and XRD analysis for sample Kis-1/36 (Fig. 4).

All minerals of the coronadite group are suitable for K-Ar and 39

Ar-40

Ar geochronological

investigations, if they contain some structural K, because they have a tunnel structure which

ensures retentiveness of argon (Vasconcelos et al., 1992, 1994, 1995; Lippolt and Hautmann,

ACC

EPTE

D M

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

9

1995). In the majority of 39

Ar-40

Ar or K-Ar studies of Mn oxides, cryptomelane refers to Mn

oxide materials which contain some K, even if other cations are dominant.

Mn oxides commonly used in 39

Ar-40

Ar studies are compact fine-grained aggregates of

needle-shaped crystals (e.g. Vasconcelos et al., 1994; Ruffet et al., 1996). These aggregates,

which are commonly laminated at macro- to microscopic scales (e.g. RGM 13933 and Kis-

1/36 in Figure 4), record successive stages of mineral deposition by accretion or

transformation (dissolution/reprecipitation), representing a range of

equilibrium/disequilibrium conditions. Development of a laminar structure is controlled by

changes in fluid composition in terms of element availability and mobility, notably for

potassium (cryptomelane formation) or aluminum (lithiophorite formation), resulting in K

content variations between laminae (Fig. 4). K content variability can also be related to

multiple interactions taking place between Mn oxides and fluids that enter the system after

initial deposition, resulting in deposits recording different phases of mineralization. As a

result, aggregates form over millions of years (1 to 5 mm/Ma for laminated concretions,

Hénocque et al., 1998), and they appear to be more complex if they crystallized early during

the evolution of this system.

3.2. Analysed samples

From historical drilling campaigns (Doyen, 1974), the secondary Mn ore is known to extend

down to ~100 m below the peneplain surface, i.e. down to ~1,000 m a.s.l. (Fig. 2). It occurs

as east-west trending subvertical beds dipping south, consisting of black massive manganese

oxides, derived from sedimentary layers with favourable lithological characteristics. Surface

occurrences are apparently always related to outcrops of these subvertical Mn-rich beds, with

in-situ Mn-oxide deposition. Occasional occurrences of thin manganese oxide veins are noted

at lower depth.

ACC

EPTE

D M

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

10

Manganese oxides in the Kisenge area deposits partly occur as impregnative material, in a

matrix derived from the garnet-bearing rhodochrosite-dominated rocks and graphitic shales

that constitute the primary ore deposits and intervening barren layers. In addition, the oxides

also occur abundantly as void-filling phases, ranging from thin massive veins to thick laminar

crusts.

Fifteen K-rich Mn oxides samples from the Kisenge deposits were selected from the mineral

and core collections of the Royal Museum for Central Africa (RMCA). The selected

specimens are part of several series of samples of the Kisenge deposits that were collected

mainly between 1945 to 1965, around the time when exploitation started (1951; Doyen,

1974). All samples used for 39

Ar-40

Ar geochronology are K-rich, with a composition close to

that of the cryptomelane K-end member of the coronadite group.

Five samples were taken from the Kis-1 oblique core (Figs 2, 5; Table 2), which crosses Mn

oxide intervals at depth, mostly from -50 to -70 m (along core) and perhaps less continuously

between -80 and -90 m. Below -90 m the sediments in the core are dominated by barren

shale, with one 3 m-thick intercalation of unaltered Mn-carbonate (rhodochrosite) deposits (-

130 m), overlying a 4m-thick interval with abundant Mn-rich garnet (spessartine). Samples

were taken from the main -50 to -70 m interval (Fig. 2), where several cryptomelane-rich

subsamples could be isolated for 39

Ar-40

Ar age determination. One sample (Kis-1/30)

represents a Mn-oxide vein crossing the deposits above the -50 to -70 m interval, at -46 m

depth.

All other analysed specimens are surface samples (Table 2; Fig. 5), again meeting the

requirement of high K content (checked by SEM-EDS).

Although the mineralogical homogeneity of the samples is unambiguously confirmed by

XRD analysis, physical heterogeneities are observed, mostly expressed as lamination (Figs.

ACC

EPTE

D M

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

11

4, 7). The origin of lamination is not precisely known: it has been noted also for malachite

(De Putter et al., 2010) and might result, for both minerals, from variations in chemical

composition and/or fabric between laminae (Mees et al., 2012). Occasional accompanying

minerals detected for the analysed features are lithiophorite (Kis-1/36, Kis-1/53, RGM

14296) and quartz (Kis-1/36, Kis-1/62). Traces of mica were only detected for RGM 17006,

which yields a non-interpretable 39

Ar-40

Ar age spectrum (see further).

4. Methods

The mineralogical composition of the specimens was determined by X-ray diffraction

analysis, with a Philips diffractometer (PW3710), using CuKα radiation (40 kV, 30 mA) and

a scanning speed of 2.5 s per 0.02° 2θ. All samples were screened using a JEOL 7500-F

scanning electron microscope at 15 kV (UNamur, Belgium).

The samples for 39

Ar-40

Ar age determination were wrapped in Al foil to form small packets

(11 11 mm) that were stacked to form columns within which fluence monitors were

inserted every 10 samples. Two distinct irradiations were performed at the McMaster reactor

(Hamilton, Canada). The first irradiation (samples KIS 1-30, KIS 1-58, RGM1769,

RGM13200 and RGM13201) used the 5C high flux location without Cd-shielding and lasted

16.667 hr (J/h ≈ 3.5x10-4

h-1

). The second irradiation which concerned all other samples

(including a second set of grains from sample RGM 13201) used a medium flux location (8E)

with Cd-shielding and lasted 52 hr (J/h ≈ 4.4x10-5

h-1

). In both cases irradiation standard was

sanidine TCRs (28.608 ± 0.033 Ma; Renne et al., 1998, 2010, 2011). Sample arrangement

during irradiation allowed monitoring of the flux gradient with a precision of 0.2 %.

Heating steps were performed with a CO2 Synrad laser (Ruffet et al., 1991, 1995). The five

argon isotopes and the background baselines were measured in eleven cycles, in peak-

ACC

EPTE

D M

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

12

jumping mode. Blanks were performed routinely each first or third/fourth run, and subtracted

from subsequent sample gas fractions. All isotopic measurements are corrected for K, Ca and

Cl isotopic interferences, mass discrimination and atmospheric argon contamination.

Apparent age errors are plotted at the 1 level and do not include errors on 40

Ar*/39

ArK ratio,

monitor age and decay constant. The latter are included in the final calculation of the

(pseudo-)plateau age error margins or for apparent ages individually cited. Analyses were

performed on a Map215

mass spectrometer.

It is commonly considered that a plateau is obtained when calculated 40

Ar*/39

ArK ratios of at

least three consecutive steps, comprising a minimum of 70 % of the 39

Ar released, agree

within 1 or 2 error bars with the weighted mean calculated 40

Ar*/39

ArK ratio of the plateau

segment. Pseudo-plateau ages (PPA) can be defined with less than 70% of the 39

Ar released.

All ages are reported at the 1 level.

Analytical data and parameters used for calculations (e.g. isotopic ratios measured on K, Ca

and Cl pure salts; mass discrimination; atmospheric argon ratios; J parameter; decay

constants) and reference sources are available in supplementary data repository.

5. Results

5.1. Age spectra characteristics

As frequently observed during 39

Ar-40

Ar geochronological studies of weathering profiles,

only the youngest samples, less likely to be affected by multiple crystallization events, yield

simple results characterized by flat age spectra allowing unambiguous plateau age calculation

(Fig. 6). This is illustrated by one of the lowest samples of the Kis-1 core (sample Kis-1/58),

with a flat age spectrum which yields a calculated plateau age of 3.6 ± 0.2 Ma. For this

ACC

EPTE

D M

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

13

sample, the c. 2.3 Ma age reported by Decrée et al. (2010) was recalculated at c. 3.6 Ma,

using updated decay constants and atmospheric argon ratios. This age is corroborated by a

plateau age of 3.6 ± 0.1 Ma yielded by surface sample RGM 1769. The youngest age

obtained by the present study is a plateau age of 2.6 ± 0.2 Ma calculated for the age spectrum

of the RGM 13200 surface sample.

However, even sample Kis-1/62, the lowest analysed Kis-1 core sample, displays a

„disturbed‟ age spectrum, resulting from mixing of the two components previously identified

(c. 2.6 and 3.6 Ma).

Most other analysed cryptomelane-rich samples are probably composite aggregates of Mn

oxides with distinct K contents which crystallized at different stages. Some components of

these analysed materials may have low K contents and the conventional age spectrum poorly

accounts for this, as it uses the percentage of total 39

ArK (# to K) degassed as discriminating

parameter, which is not weighted by temperature (or laser power) increment. In this case,

some alternative or supplementary types of representation or calculation can be used, such as

(i) correlation (36

Ar/40

Ar vs. 39

ArK/40

Ar* inverse isochron; Turner, 1971; Roddick et al.,

1980; Hanes et al., 1985); (ii) pseudo-plateau age (PPA) calculations, i.e. plateau age

calculated with less than 70% of the 39

ArK released; (iii) degassing analysis, i.e. visualization

of degassing peak(s) ((xAr/T°)/(

xAr/T°)Max versus %

39ArK, with x=36 to 40) or the use of

weighted age spectra (apparent ages versus %(39

ArK/T°)/(39

ArK/T°)Max).

The usefulness of such alternative procedures can be illustrated for the results obtained for

sample RGM 10727c (Fig. 7). The staircase shape of the conventional age spectrum (Fig.

7C) suggests that the analysed material was composite. The regular shape of the spectrum

could be explained by mixing of only two distinct phases, whereby a maximum age for the

youngest phase () would be estimated from low-temperature apparent ages, with a PPA of

ACC

EPTE

D M

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

14

c. 10.8 Ma, whereas a minimum age estimate for the oldest phase () would be suggested by

high temperature apparent ages, with a PPA of c. 14.2 Ma. In fact, a detailed analysis of the

data shows that there are at least three distinct phases within the analysed grain. The

degassing diagrams (40

Ar* and 36

Ar, Fig. 7A) display a peak in the low to intermediate

temperature steps, with a maximum at c. 15% of 39

ArK degassing, which irrefutably

evidences a „hidden‟ phase (), poorly expressed in the conventional age spectrum (PPA

13.6 ± 0.1 Ma) but magnified when taking into account degassing intensity of 39

ArK

(%[39

ArK/T°]) (Fig. 7D). This alternative representation of apparent ages clearly shows two

successive staircase segments (I and II, Fig. 7D), which express transitions between the three

distinct phases constituting the analysed grain. The existence of these three distinct phases is

furthermore confirmed by inverse isochron analysis (Fig. 7B), which yields three distinct

isochrons resulting from mixing between an atmospheric reservoir, attested by three

concordant (40

Ar/36

Ar)i atmospheric ratios, and three distinct radiogenic components. These

isochrons are joined by transition steps (I and II, Fig. 7D), which are the counterparts of the

age spectra PPA connections (Fig. 7C,D). Obviously, isochron ages are fully concordant with

calculated pseudo-plateau ages.

The youngest phase (, 10.8 ± 0.2 Ma) is weakly expressed in the conventional and

alternative age spectra (Fig. 7C,D), either as a result of its low total volume within the

analysed grain or more probably because of its low K content. Nevertheless, its imprint is

strong in the inverse isochron diagram (Fig. 7B), due to its related high atmospheric

contamination (93.4 % to 68.1 %) with a clearly defined and specific 36

Aratm degassing peak

(Fig. 7A), and its existence is established with certainty.

A striking feature of sample RGM 10727c is the progressiveness of degassing of the distinct

phases (→ → ) constituting the analysed composite grain, from youngest to oldest. It

ACC

EPTE

D M

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

15

could be argued that this feature is linked to heterogeneity of the analysed material (Fig. 7E),

degassing first the laminated cover deposit, corresponding to the youngest phase (and then

the rounded aggregates of the covered substrate, corresponding to the older phases (→ ).

Nevertheless, because all identified phases were simultaneously exposed to the laser beam

during step-heating (see picture in Figure 7E, taken before the grain fusion step, with surface

exposed to laser beam perpendicular to lamination in cross-section), the spreading of

degassing peaks must be related to differences in 40

Ar* retentiveness, whereby values

decrease as younger phases are formed (→ → ). This could be attributed to differences

in crystal structure properties. The retentiveness variability of successive crystallized phases

could be controlled by K availability, which generally decreases with the lowering of the

weathering front; alternatively, it could also change through progressive increase in

crystallinity, with „ageing‟ of the deposits.

Most of the experiments performed on samples from this study evidence younger

superimposed phases in the low to intermediate 39

ArK degassing domain which usually can be

characterized through PPA and inverse isochron calculations. On the other hand, preservation

of the older main phase depends on the sequential occurrence of degassing of the various Mn

oxides generations. Degassing of the younger generation can overlap with or be

superimposed on that of the older main generation. Such configurations generate fully

staircase-shaped age spectra, without flat segments in the high temperature domain of the

spectra (e.g. RGM 10739 and 14296, Fig. 6), or hump-shaped patterns (Kis-1/30, 1/36 and

1/53, RGM 1767c and 10724, Fig. 6). The highest apparent age reached in a staircase-shaped

age spectrum would be a minimum estimate of the age of the main phase. Similar conclusions

could be drawn for hump-shaped age spectra but these are more delicate to interpret.

ACC

EPTE

D M

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

16

5.2. 39

Ar-40

Ar ages

In view of the preceding discussion, the oldest age obtained in this study (19.2 ± 0.1 Ma,

RGM 1767c) is probably a minimum estimate for the age of the oldest phase, which is

probably also the case for the age of 17.6 ± 0.1 Ma obtained for sample RGM 10724. On the

other hand, the age of 10.5 ± 0.1 Ma yielded by sample Kis-1/30 (Fig. 6) is probably the best

estimate for the age of the oldest phase recorded by core Kis-1, coherent with its position in

the weathering profile (Fig. 2). A c. 10.5 Ma phase is recurrently observed for other samples,

either as remnants of a „main‟ phase (10.5 ± 0.1 Ma for RGM 10739, Fig. 6), or as a low-

temperature superimposed phase (10.8 ± 0.2 Ma for RGM 10727c, 10.6 ± 0.4 Ma for RGM

13933). It is also clearly evidenced through isochrones analyses, yielding 10.8 ± 0.3 Ma for

RGM 10727c (Fig. 7B) and 10.7 ± 1.1 Ma for RGM 13933 (Fig. 6). Similarly, a c. 2.6 Ma

episode is clearly recognized for the weathering sequence. As previously discussed, it is

observed as a main component (RGM 13200, with the best plateau obtained during this study

at 2.6 ± 0.2 Ma; Fig. 6), but also as a low-temperature superimposed phase, both in the core

(Kis-1/62, 2.7 ± 0.2 Ma) and in surface samples (RGM 14296, 2.6 ± 0.2 Ma). The 2.6 Ma

phase is relatively poorly expressed in the age spectra, but it is more apparent through

isochron analysis (2.7 ± 0.3 Ma for Kis-1/62, 2.6 ± 0.4 Ma for RGM 14296; Fig. 8).

Surface samples yield a broader age range than the core samples, with an oldest phase of at

least c. 19.2 Ma and a youngest phase of c. 2.6 Ma. The three main components identified for

core Kis-1 (10.5 Ma; 3.6 Ma; 2.6 Ma) can also be observed for these samples as main and

superimposed phases.

As expected, the older phases are the least well expressed (e.g. Vasconcelos and Conroy,

2003). An ≥19.2 ± 0.1 Ma age is certainly recorded for sample RGM 1767c and possibly for

RGM 10724. On the other hand, a c. 15.7 Ma phase is rather well preserved, as it is observed

as a main component for samples RGM 13933 (15.7 ± 0.1 Ma) and possibly RGM 14296

ACC

EPTE

D M

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

17

(>15.0 ± 0.1 Ma), and also as one of both superimposed components for sample RGM 1767c

(15.9 ± 0.1 Ma). In the same way, a parallel could be drawn between the c. 14.2 Ma

component derived from the previously presented analysis for sample RGM 10727c, and the

lowest-temperature subordinate c. 14.9 Ma component observed for sample RGM 1767c.

Finally, a weakly expressed c. 13.6 Ma phase, detected as a superimposed component for

sample RGM 10727c (see above) is also discernible as a low-temperature component through

isochron analysis for sample RGM 10724 but never as main phase. Any attempt to identify

others phases would be highly speculative because the available data do not allow cross-

checking within those other parts of the age range.

In summary, three main phases are recognized both for the Kis-1 core and for surface

samples, corresponding to c. 10.5 Ma, 3.6 Ma and 2.6 Ma. In addition, analyses of surface

samples show that cryptomelane also formed before c. 19.2 Ma and sporadically during at

least three periods in the course of the middle Miocene, at c. 15.7 Ma, c. 14.2 Ma and

c. 13.6 Ma., before the c. 10.5 Ma main phase.

6. Discussion

6.1. Significance of the obtained ages

39Ar-

40Ar dating results obtained for the Kisenge Mn oxide deposits suggest that the area

occupied an overall stable position across the Mio-Pliocene time interval (c. 20 to c. 2.5 Ma).

The results are marked by a lack of Paleocene or Eocene ages, which have been widely

recorded for similar paleosurface-related deposits occurring in other parts of Africa and in

South America (Vasconcelos et al., 1994; Ruffet et al., 1996). Scattered but widespread

occurrences of bauxite throughout the DR Congo have been attributed to a major Paleogene

lateritization episode that affected the whole of Central Africa (Guillocheau et al., 2015), as

also recorded for West Africa (Tardy et al., 1991; Chardon et al., 2006; Burke & Gunnell,

ACC

EPTE

D M

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

18

2008; Butt and Bristow, 2012). The present study indicates that, in Central Africa, these

deposits have subsequently been eroded, most likely during the first major uplift stage of the

East African Plateau (Dome), during the Eocene-Oligocene (Ebinger et al., 1993; Wichura et

al., 2010), leaving no traces of this major, possibly continent-wide, lateritization event in the

Katanga plateau area.

The oldest recorded ages for the Kisenge deposits (≥ 19.2 Ma, 15.7 Ma, 14.2 Ma, 13.6 Ma)

do not correspond with any known periods of regional low tectonic activity which would

have favoured landscape stability and duricrust development. A next and major phase of Mn

oxide formation (10.5-11 Ma) roughly coincides with a period of EAP stability during the

middle to upper Miocene. No significant events are subsequently recorded for the upper

Miocene, when active rifting took place in the East and high uplift rates were attained in the

Congo Basin. Uplift rates for this region were much lower during the Pliocene, a period

marked at Kisenge by well-expressed Mn oxide occurrences, around 3.6 Ma and 2.6 Ma. The

latter roughly coincides with start of the Pleistocene aridity in the EARS and EAP areas

(Goudie, 2005; Gasse, 2006; Bonnefille, 2010). This appears to be mainly significant in terms

of a lack of more recent recorded stages of Mn oxide accumulation, as the preceding upper

Miocene to Pliocene wetter phase (Senut et al., 2009) may have presented more favourable

conditions. Secondary ore deposition in humid conditions in Katanga is suggested by the

rather common occurrence of malachite in the form of stalactitic speleothems (Decrée et al.,

2010; De Putter et al., 2010), a morphology that is also known for Mn oxide deposits from

Kisenge (RMCA collection). The preservation of the 2.6 Ma deposit also implies that they

were not entirely removed by erosion following their formation, despite Quaternary

landscape evolution (De Dapper, 1991).

Overall, the Miocene and Pliocene weathering events identified in this study fit fairly well

with periods of laterite formation recognized for the African continent (Hénocque et al.,

ACC

EPTE

D M

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

19

1998; van Niekerk et al., 1999; Colin et al., 2005; Boni et al., 2007; Beauvais et al., 2008;

Decrée et al., 2010; Gutzmer et al., 2012 ).

6.2. Implications for paleosurface evolution

The uplift rate computed for the Congo basin based on sediment discharge data for the

offshore Congo fan amounts to ~5m/My within the 11-5 Ma time interval. In the 5-0 Ma

interval, the uplift slowed down to the West of the study area (Congo basin) but it increased

to the East, in the EAP area (Fig. 3; Lavier et al., 2001; Guillocheau et al., 2015).

The three phases dated at c. 10.5 Ma, 3.6 Ma and 2.6 Ma by the 39

Ar-40

Ar study of the Kis-1

core samples are major episodes in the weathering history of the area (Fig. 9). The 10.5 Ma

event follows several older ones (>19.2 to 13.6 Ma). During the 10.5 Ma stage, the

weathering front descended to 1034m a.s.l., below a surface level whose exact elevation at

the time is not known. This was followed by a c. 6.9 Myr quiescent period. At c. 3.6 Ma, new

meteoric fluid input affected the whole preexisting weathering column and allows a next

phase of weathering front lowering, down to 1026m a.s.l. For the subsequent phase, at

c. 2.6 Ma, there is no evidence for further lowering of the weathering front. Minor (undated)

Mn oxide occurrences are however observed at ~1015m a.s.l., and the unweathered Mn-

carbonate ore is cut at ~980m a.s.l. (Figs 2, 9). The 39

Ar-40

Ar dating results for the Kis-1

core samples, taken at known depth and present-day absolute altitude, show that a total

thickness of ≥100 m (1130-1026 m a.s.l.) has been affected by the downward penetration of

the weathering front, over a 6.9 My period (from 10.5 to 3.6 Ma) (Figs 2, 9). This figure has

to be considered with caution, as the downward penetration of the weathering front might be

irregular in time and space, and will vary with host-rock permeability and porosity.

ACC

EPTE

D M

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

20

Similar features were recognized for the manganese ore deposit of Tambao in northern

Burkina Faso with more limited thickness (c. 70-80 m) of the main Mn-oxide bodies, and

occasional occurrences of manganese oxide veins down to c. 110 m beneath hilltop level

(Beauvais et al. 2008). Detailed correlations between 39

Ar-40

Ar cryptomelane ages and

geomorphologic features have been performed in the Tambao area, where five Cenozoic

paleosurfaces have been recognized and dated (Beauvais et al., 2008). They were

subsequently used as a basis for a regional analysis for West Africa (Beauvais and Chardon,

2013). Such precise correlations are not yet possible for the Kisenge area where critical

information regarding denudation rates, uplift rates, and recent tectonic history are lacking.

Paleosurfaces were however recognized to the East, in the Copperbelt (De Dapper, 1991),

and some might be derived from a DEM for the study area (Fig. 9), but it seems premature to

equate them with the 39

Ar-40

Ar cryptomelane ages obtained in the Kis-1 core.

The thickness of the weathering zone at Kisenge (≥100 m, formed over ~7 My) seems small

when compared to the western part of the Copperbelt, where the substrate is weathered down

to ~300 meters below the surface (Dewaele et al., 2006). Deeper weathering in the

Copperbelt was probably favoured by the calcareous facies that hosts the Cu-Co

mineralization (De Putter et al., 2010). By contrast, the rhodochrosite ore of Kisenge is

interlayered within siliciclastic sediments (shale), less prone to dissolution and the formation

of caves and other karstic cavities. The 15 m/My rate in Kisenge is however significantly

higher than the ~4m/My value calculated for poorly developed weathering profiles in

tectonically quiescent zones in the Proterozoic Mount Isa Block in Northern Australia

(Vasconcelos and Conroy, 2003). The difference between the two rates may be seen as a

further indication that the study area was (and still is) a tectonically active zone in the period

considered in this study (Kipata et al., 2013), with discrete quiescent phases allowing the

formation of Mn oxides.

ACC

EPTE

D M

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

21

6.3. Comparison with other African supergene deposits

Radiometric ages for supergene manganese deposits have been published for several parts of

Africa, which can be grouped as West Africa and southern Africa (Table 1). Palaeocene ages

were only obtained for West Africa (Colin et al., 2005; Beauvais et al., 2008) whereas

Eocene ages are recorded for both West and southern Africa (Hénocque et al., 1998; Colin et

al., 2005; Beauvais et al., 2008; Gutzmer et al., 2012). As mentioned above, any possible

occurrences of these generations of duricrusts in Kisenge have been eroded during a major

Eocene-Oligocene uplift stage.

The pre-19 Ma phase that is tentatively recognized for Kisenge may correspond to the

Oligocene phase (c. 25 Ma) that is documented for West Africa (Colin et al., 2005; Beauvais

et al., 2008) and southern Africa (Boni et al., 2007; Gutzmer et al., 2012).

For the Mio-Pliocene, the Kisenge deposits record ages that are not systematically found

elsewhere in Africa. The 10.5-11 Ma event only corresponds to a roughly simultaneous event

in the Kalahari Manganese Field, South Africa (Gutzmer et al., 2012). This correspondence

suggests that the relative stable tectonic setting experienced in the Katanga area extended far

southward at that stage. A common evolution of both areas in the Neogene is further

demonstrated by the deposition of a sandy sedimentary cover, as part of the so-called

„Kalahari Group‟, on a vast area stretching from the southern margin of the Congo Basin

(including Katanga) to the Kalahari area in southern Africa (Linol et al., 2015b).

The rest of the Katanga paleosurface record differs from records for other parts of Africa. The

interpretation of this difference is not straightforward, in view of possible sampling bias, the

lack of regional geomorphological studies and the likely influence of Oligocene EARS

opening.

ACC

EPTE

D M

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

22

From the beginning of the Paleogene, the geodynamic history of the region is largely

controlled by the evolution of the neighbouring Eastern African Plateau. Uplift, erosion, and

the opening of the EARS at c. 25 Ma are clearly events whose effects outweigh those of

processes responsible for the development of continent-wide paleosurfaces. This major

tectonic forcing has even had a strong impact on continental ocean-atmosphere circulation,

and hence on the climate and precipitation regimes prevailing in Central Africa (Prömmel et

al., 2013).

7. Conclusion

The 39

Ar-40

Ar geochronological study of cryptomelane of the Kisenge manganese deposit

provides information on the Mio-Pliocene evolution of an area that is located South of the

Congo Basin and West of the EARS. Discrete phases of cryptomelane formation are dated at

c. 10.5 Ma, 3.6 Ma and 2.6 Ma for an interval crossed by the Kis-1 core. Dating of core

samples dating also suggests that a total thickness of ≥100 m has been affected by the

downward penetration of the weathering front, over a 6.9 My period (from 10.5 to 3.6 Ma).

This value seems low when compared to the total thickness of weathered sediments in the

Copperbelt which at least locally sometimes exceeds 300m. The difference may be due to the

abundance of calcareous host rocks in the Copperbelt area, susceptible to dissolution (De

Putter et al., 2010).

In addition to the < 10.5 Ma ages obtained for the Kis-1 core, surface samples from the

Kisenge area also record a ≥ c. 19.2 Ma phase, as well as c. 15.7 Ma, 14.2 Ma and 13.6 Ma

phases. The absence of Eocene and Oligocene ages at Kisenge most probably results from the

removal of corresponding generations of duricrusts during major Paleogene uplift stages. As

a result, the Paleogene evolution of the study area cannot be compared with that of other parts

ACC

EPTE

D M

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

23

of Africa for which supergene manganese deposits have been studied (Table 1), and a series

of middle Miocene events (18-11.5 Ma) recognized (Van Niekerk et al., 1999; Beauvais et al,

2008; Gutzmer et al., 2012). On the other hand, the c. 10.5-11 Ma event matches a roughly

simultaneous event documented for the Kalahari Manganese Field (Gutzmer et al., 2012), but

not for West Africa. The rest of the Katanga paleosurface record is hardly comparable with

records for other parts of Africa. The difference is most probably related to the specific

regional geodynamic setting of the study area, which is influenced along its western and

northern margins by Congo Basin dynamics and along its eastern border by the uplift of the

East African Plateau and the opening of the EARS, from c. 25 Ma onward (Roberts et al.,

2012). Such major tectonic events obviously prevail over the development of continent-wide

paleosurfaces, and may also result in significant climate changes over large parts of Central

Africa (Prömmel et al., 2013).

The present study provides a framework for the age of secondary ore deposits of the EARS

region, both supergene and eluvial. These deposits include the major supergene Cu-Co ores

of the Copperbelt region, to which the results of a preliminary 39

Ar-40

Ar study of the Kisenge

deposits have already been applied (Decrée et al., 2010). Other examples are gold and tin-

tantalum-tungsten eluvial deposits in the Great Lakes region, for which age information is

still lacking, and the Nb-REE deposit at Lueshe, Kivu, for which an undated supergene

imprint has been identified (Nasraoui et al., 2000). This paper highlights the fact that the

study of Cenozoic morphotectonic evolution of large parts of Central Africa, and its link with

the formation of supergene ore deposits in this area, are still in relative infancy.

ACC

EPTE

D M

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

24

Acknowledgements

This study is a contribution the Paleurafrica research project (BR/121/A3/Paleurafrica) of the

Belgian Science Policy Office (BELSPO). C. Charlier (Service de Microscopie Electronique)

is thanked for his help in using the SEM facility at the Université de Namur (Belgium). Dr M.

Laghmouch (Royal Museum for Central Africa) has provided the DEM data and cartographic

material used in this paper. Two anonymous reviewers are thanked for their constructive

comments.

References

Alexandre J., 2002. Les cuirasses latéritiques et autres formations ferrugineuses tropicales –

Exemple du Haut-Katanga méridional. Annales du Musée Royal de l‟Afrique Centrale,

Sciences géologiques 107, 118p.

Anka, Z., Séranne, M., Lopez, M., Scheck-Wenderoth, M., Savoye, B., 2009. The long-term

evolution of the Congo deep-sea fan: A basin-wide view of the interaction between a giant

submarine fan and a mature passive margin (ZaiAngo project). Tectonophysics 470, 42–56.

Anka, Z., Séranne, M., di Primio, R., 2010. Evidence of a large upper Cretaceous depocentre

across the Continent-Ocean boundary of the Congo-Angola basin. Implications for palaeo-

drainage and potential ultra-deep source rocks. Marine and Petroleum Geology 27, 301-611.

Beauvais, A., Chardon, D., 2013. Modes, tempo, and spatial variability of Cenozoic cratonic

denudation: the West African example. Geochemistry, Geophysics, Geosystems 14, 1590-

1608.

ACC

EPTE

D M

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

25

Beauvais, A., Ruffet, G., Hénocque, O., Colin, F., 2008. Chemical and physical erosion

rhythms of the West African Cenozoic morphogenesis: The 39

Ar-40

Ar dating of supergene K-

Mn oxides. Journal of Geophysical Research 113, F04007, doi:10.1029/2008JF000996

Beugnies, A., 1954. La nappe phréatique des environs d‟Élisabethville et les phénomènes

connexes d‟altération superficielle. Publications du Comité Spécial du Katanga (C.S.K.),

série A, fascicule 3, Annales du Service des Mines et du Service Géographique et Géologique

17, 3-54.

Biagioni, C., Capalbo, C., Pasero M., 2013. Nomenclature tunings in the hollandite

supergroup. European Journal of Mineralogy 25, 85-90.

Boni, M., Terracciano, R., Evans, N.J., Laukamp, C., Schneider, J., Bechstädt, T., 2007.

Genesis of Vanadium Ores in the Otavi Mountainland, Namibia. Economic Geology 102,

441–469.

Bonnefille, R., 2010. Cenozoic vegetation, climate changes and hominid evolution in tropical

Africa. Global and Planetary Change 72, 390-411.

Burke, K., Gunnell, Y., 2008. The African erosion surface: a continental-scale synthesis of

geomorphology, tectonics, and environmental change over the past 180 million years.

Geological Society of America, Memoir 201, 66p.

Butt, C.R.M, Bristow, A.P.J., 2012. Relief inversion in the geomorphological evolution of

sub-Saharan West Africa. Geomorphology 185, 16-26.

Cahen, L., Snelling, N.J., 1984. The geochronology and evolution of Africa. Clarendon,

Oxford, 512p.

ACC

EPTE

D M

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

26

Cailteux, J.L.H., Kampunzu, A.B., Lerouge, C., Kaputo, A.K., Milesi, J.P., 2005. Genesis of

sediment-hosted stratiform copper-cobalt deposits, Central African Copperbelt. Journal of

African Earth Sciences 42, 134-158.

Chardon, D., Chevillotte, V., Beauvais, A., Grandin, G., Boulangé, B., 2006. Planation,

bauxites and epeirogeny : one or two paleosurfaces on the West African margin ?

Geomorphology 82, 273-282.

Chisonga, B.C., Gutzmer, J., Beukes, N.K., Huizenga, J.M., 2012. Nature and origin of the

protolith succession to the Paleoproterozoic Serra do Navio manganese deposit, Amapa

Province, Brazil. Ore Geology Reviews 47, 59-76.

Colin, F., Beauvais, A., Ruffet, G., Hénocque, O., 2005. First 39

Ar-40

Ar geochronology of

lateritic manganiferous pisolites : implications for the Palaeogene history of a West African

landscape. Earth and Planetary Science Letters 238, 172-188.

Decrée S., Deloule E., Ruffet G., Dewaele S., Mees F., Marignac C., Yans J., De Putter Th.,

2010. Geodynamics and climate controls in the formation of Mio-Pliocene world-class

oxidized cobalt and manganese ores in the Katanga province, DR Congo. Mineralium

Deposita 45, 621-629.

Dewaele, S., Muchez, P., Vets, J., Fernandez-Alonzo, M., Tack, L., 2006. Multiphase origin

of the Cu–Co ore deposits in the western part of the Lufilian fold-and-thrust belt, Katanga

(Democratic Republic of Congo). Journal of African Earth Sciences 46, 455–469.

De Dapper, M., 1991. Late Quaternary geomorphological evolution of the sand-covered

plateaus near Kolwezi, Southern Shaba, Zaïre. Bulletin de la Société géographique de Liège

27, 157-173.

ACC

EPTE

D M

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

27

De Putter Th., Mees F., Decrée S., Dewaele S., 2010. Malachite, an indicator of major

Pliocene Cu remobilization in a karstic environment (Katanga, Democratic Republic of

Congo). Ore Geology Reviews 38, 90-100.

Dill, H.G., Weber, B., Botz, R., 2013. Metalliferous duricrusts (“orecretes”) – markers of

weathering: A mineralogical and climatic-geomorphological approach to supergene Pb-Zn-

Cu-Sb-P mineralization on different parent materials. Neues Jahrbuch für Mineralogie

Abhandlungen 190, 123-195.

Doyen, L., 1974. Etude métallogénique des gisements de manganèse de Kisenge – Kamata –

Kapolo. Unpubl. PhD thesis Brussels University, 493p.

Ebinger, C.J., Yemane, T., Woldegabriel, G., Aronson, J.L., Walter, R.C., 1993. Late

Eocene–Recent volcanism and faulting in the southern main Ethiopian rift. Journal of the

Geological Society 150, 99-108.

Fernandez-Alonso, M., Cutten, H., De Waele, B., Tack, L., Tahon, A., Baudet, D., Barritt,

S.D., 2012. The Mesoproterozoic Karagwe-Ankole Belt (formerly the NE Kibara Belt): The

result of prolonged extensional intracratonic basin development punctuated by two short-

lived far-field compressional events. Precambrian Research 216-219, 63- 86.

Gasse, F., 2006. Climate and hydrological changes in tropical Africa during the past million

years. Comptes-Rendus Palevol 5, 35-43.

Goudie, A.S., 2005. The drainage of Africa since the Cretaceous. Geomorphology 67, 437-

456.

Guillocheau, F., Chelalou, R., Linol, B., Dauteuil, O., Robin, C., Mvondo, F., Callec, Y.,

Colin, J.-P., 2015. Cenozoic landscape evolution in and around the Congo Basin : constraints

ACC

EPTE

D M

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

28

from sediments and planation surfaces. In M.J. de Wit, F. Guillocheau, M.C.J. de Wit (eds),

Geology and resource potential of the Congo Basin, 271-314.

Gutzmer, J., Du Plooy, A.P., Beukes, N.J., 2012. Timing of supergene enrichment of low-

grade sedimentary manganese ores in the Kalahari Manganese Field, South Africa. Ore

Geology Reviews 47, 136-153.

Hanes, J.A., York, D., Hall, C.M., 1985. An 40

Ar/39

Ar geochronological and electron

microprobe investigation of an Archean pyroxenite and its bearing on ancient atmospheric

compositions. Canadian Journal of Earth Science 22, 947-958.

Hénocque, O., Ruffet, G., Colin, F., Féraud, G., 1998. 40

Ar/39

Ar dating of West African

lateritic cryptomelanes. Geochimica et Cosmochimica Acta 62(16), 2739-2756.

Hitzman, M.W., Broughton, D., Selley, D., Woodhead, J., Wood, D., Bull, S., 2012. The

Central African Copperbelt: diverse stratigraphic, structural, and temporal settings in the

world‟s largest sedimentary copper district. Society of Economic Geologists, Special

Publication 16, 487-514.

Kipata, M.L., Delvaux, D., Sebagenzi, M.N., Cailteux, J., Sintubin, M., 2013. Brittle tectonic

and stress field evolution in the Pan-African Lufilian arc and its foreland (Katanga, DRC):

from orogenic compression to extensional collapse, transpressional inversion and transition to

rifting. Geologica Belgica 16, 1-17.

Kuleshov, V.N., 2011. Manganese Deposits: Communication 2. Major epochs and phases of

manganese accumulation in the Earth‟s history. Lithology and Mineral Resources 46, 546-

565.

Lavier, L.L., Steckler, M.S., Brigaud, F., 2001. Climatic and tectonic control on the Cenozoic

evolution of the West African margin. Marine Geology 178, 63-80.

ACC

EPTE

D M

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

29

Ledent, D., Lay, C. Delhal, J., 1962. Premières données sur l‟âge absolu des formations

anciennes du « socle du Kasaï » (Congo méridional). Bulletin de la Société belge de

Géologie, XXI, 223-235.

Lepersonne, J.,1974. Carte géologique du Zaïre. Musée Royal de l‟Afrique Centrale,

Tervuren.

Linol, B., de Wit, M.J., Guillocheau, F., Robin, C., Dauteuil, O., 2015a. Multiphase

Phanerozoic subsidence and uplift history recorded in the Congo Basin : a complex successor

basin. In M.J. de Wit et al. (eds), Geology and resource potential of the Congo Basin.

Regional Geology Reviews, 213-227.

Linol, B., de Wit, M.J., Guillocheau, F., de Wit, M.C.J., Anka, Z., Colin, J.-P., 2015b.

Formation and collapse of the Kalahari duricrust [„African Surface‟] across the Congo Basin,

with implications for changes in rates of Cenozoic off-shore sedimentation. In M.J. de Wit et

al. (eds), Geology and resource potential of the Congo Basin. Regional Geology Reviews,

193-210.

Lippolt, H.J., Hautmann, S., 1995. 40

Ar/39

Ar ages of Precambrian manganese ore minerals

from Sweden, India and Morocco, Mineralium Deposita 30, 246-256.

Marchandise, H., 1958. Le gisement et les minerais de manganèse de Kisenge (Congo belge).

Bulletin de la Société Belge de Géologie, 67 : 187-210.

Mees, F., De Putter, Th., Decrée, S., Dewaele, S., 2012. Petrographical features of malachite

from Katanga as indicators of mineral formation processes – preliminary results. In: S.

Decrée, Th. De Putter (eds), Le secteur minier de la République démocratique du Congo à la

croisée des chemins. Royal Museum for Central Africa, Tervuren, p. 73-74.

ACC

EPTE

D M

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

30

Nasraoui, M., Toulkeridis, T., Clauer, N., Bilal, E., 2000. Differentiated hydrothermal and

meteoric alterations of the Lueshe carbonatite complex (Democratic Republic of Congo)

identified by a REE study combined with a sequential acid-leaching experiment. Chemical

Geology 165, 109-132.

Nyame, F.K., 2008. Petrography and geochemistry of intraclastic manganese-carbonates from

the ~2.2 Ga Nsuta deposit of Ghana: significance for manganese sedimentation in the

Paleoproterozoic of West Africa. Journal of African Earth Sciences 50, 133-147.

Pack, A., Gutzmer, J., Beukes, N.J., van Niekerk, H.S., Hoernes, S., 2000. Supergene

ferromanganese wad deposits derived from Permian Karoo strata along the Late Cretaceous–

Mid-Tertiary African land surface, Ryedale, South Africa. Economic Geology 95, 203-220.

Pasteels, P., Villeneuve, M., De Paepe, P., Klerkx, J., 1989. Timing of the volcanism of the

southern Kivu Province : implications for the evolution of the western branch of the East

African Rift system. Earth and Planetary Science Letters 94, 353-363.

Peters, C.R., O‟Brien, E., 1999. Palaeo-lake Congo: implications for Africa‟s late Cenozoic

climate – some unanswered questions. Proceedings of the XVth

INQUA Conference, Durban,

South Africa, 3-11 August 1999. Palaeoecology of Africa and Surrounding islands 27, 11-18.

Pik, R., Marty, B., Carignan, J., Yirgu, G., Ayalew, T., 2008. Timing of East African Rift

development in Southern Ethiopia: implication for mantle plume activity and evolution of

topography. Geology 36 (2), 167-170.

Polinard, E., 1932. Esquisse géologique de la région située au sud du parallèle de Sandoa-

Kafakumba. Annales de la Société Géologique de Belgique 54, 100-105.

ACC

EPTE

D M

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

31

Prömmel, K., Cubasch, U., Kaspar, F., 2013. A regional climate model study of the impact of

tectonic and orbital forcing on African precipitation and vegetation. Palaeogeography,

Palaeoclimatology, Palaeoecology 369, 154-162.

Renne, P.R., Balco, G., Ludwig, R.L., Mundil, R., Min, K., 2011. Response to the comment

by W.H. Schwarz et al. on "Joint determination of 40

K decay constants and 40

Ar∗/40K for the

Fish Canyon sanidine standard, and improved accuracy for 40

Ar/39

Ar geochronology" by PR

Renne et al. (2010). Geochimica et Cosmochimica Acta 75, 5097-5100.

Renne, P.R., Mundil, R., Balco, G., Min, K., Ludwi, R.L., 2011. Joint determination of 40

K

decay constants and 40

Ar∗/40K for the Fish Canyon sanidine standard, and improved accuracy

for 40

Ar/39

Ar geochronology. Geochimica et Cosmochimica Acta 74, 5349–5367.

Renne, P. R., Swisher, C. C., Deino, A. L., Karner, D. B., Owens, T. L., DePaolo, D. J., 1998.

Intercalibration of standards, absolute ages and uncertainties in 40

Ar/39

Ar dating. Chemical

Geology 145, 117-152.

Roberts, E.M., Stevens, N.J., O‟Connor, P.M., Dirks, P.H.G.M., Gottfried, M.D., Clyde,

W.C., Armstrong, R.A., Kemp, A.I.S., Hemming, S., 2012. Initiation of the western branch

of the East African Rift coeval with the eastern branch. Nature Geoscience 5, 289-294.

Roddick, J. C., Cliff, R. A., ,Rex D. C., 1980. The evolution of excess argon in alpine

biotites - A 40

Ar-39

Ar analysis. Earth Planetary Science Letters 48, 185-208.

Roy, S., 2006. Sedimentary manganese metallogenesis in response to the evolution of the

Earth system. Earth-Science Reviews 77, 273-305.

Ruffet, G., Féraud, G., Amouric, M., 1991. Comparison of 40

Ar/39

Ar conventional and laser

dating of biotites fram the North Trégor Batholith. Geochimica et Cosmochimica Acta 55,

1675–1688.

ACC

EPTE

D M

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

32

Ruffet, G., Féraud, G., Ballèvre, M., Kiénast, J.R., , 1995. Plateau ages and excess argon on

phengites: a 40

Ar/39

Ar laser probe study of alpine micas (Sesia zone). Chemical Geology 121,

327–343.

Ruffet, G., Innocent, C., Michard, A., Féraud, G., Beauvais, A., Nahon, D., Hamelin, B.,

1996. A geochronological 40

Ar/39

Ar and 87

Rb/87

Sr study of K-Mn oxides from the weathering

sequence of Azul, Brazil. Geochimica et Cosmochimica Acta 60, 2219-2232.

Schneiderhan, E.A., Gutzmer, J., Strauss, H., Mezger, K., Beukes, N.J., 2006. The

chemostratigraphy of a Paleoproterozoic MnF-BIF succession – the Voëlwater Subgroup of

the Transvaal Supergroup in Griqualand West, South Africa. South African Journal of

Geology 109, 63-80.

Schuiling, H., Grosemans, P., 1956. Les gisements de manganèse du Congo belge. Congrès

Géologique International, XXe session. Symposium sobre yacimientos de manganesos, II,

Africa : 131-142.

Senut, B., Pickford, M., Ségalen, L., 2009. Neogene desertification of Africa. Comptes-

Rendus Géosciences 341, 591-602.

Tardy, Y., Kobilsek, B., Paquet, H., 1991. Mineralogical composition and geographical

distribution of African and Brazilian periatlantic laterites. The influence of continental drift

and tropical paleoclimates during the past 150 million years and implications for India and

Australia. Journal of African Earth Sciences 12, 283-295.

Tardy, Y., Roquin, C., 1998. Dérive des continents, Paléoclimats et altérations tropicales.

BRGM, 473p.

Turner, G., 1971. 40

Ar-39

Ar ages from the lunar Maria. Earth Planetary Science Letters 11,

169-191.

ACC

EPTE

D M

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

33

USGS, 2009. Cobalt. Mineral Commodity Summaries, 195.

van Niekerk, H.S., Gutzmer, J., Beukes, N.J., Philips, D., Kiviets, G.B., 1999. An 40

Ar/39

Ar

age of supergene K-Mn oxyhydroxides in a post-Gondwana soil profile on the Highveld of

South Africa. South African Journal of Science 95, 450-454.

Vasconcelos, P.M., 1999. K-Ar and 40

Ar-39

Ar geochronology of weathering processes.

Annual Review of Earth and Planetary Sciences 27, 183-229.

Vasconcelos P.M., T.A. Becker, P.R. Renne, Brimhall G.H., 1992. Age and duration of

weathering by 40

K-40

Ar and 40

Ar/39

Ar analysis of Potassium-Manganese Oxides, Science 58,

451-455.

Vasconcelos P.M., P.R. Renne, T.A. Becker, Wenk, H.R., 1995. Mechanisms and kinetics of

atmospheric, radiogenic, and nucleogenic argon release from cryptomelane during 40

Ar/39

Ar

analysis, Geochimica Cosmochimica Acta 59 (10), 2057-2070.

Vasconcelos P.M., P.R. Renne, G.H. Brimhall, Becker, T.A., 1994. Direct dating of

weathering phenomena by 40

Ar/39

Ar and K-Ar analysis of supergene K-Mn oxides,

Geochimica Cosmochimica Acta 58, 1635-1665.

Vasconcelos, P.M., Conroy, M., 2003. Geochronology of weathering and landscape

evolution, Dugald River valley, NW Queensland, Australia. Geochimica et Cosmochimica

Acta 67, 2913-2930.

Wichura, H., Bousquet, R., Oberhänsli, R., Strecker, M.R., Trauth, M.H., 2010. Evidence for

middle Miocene uplift of the East African Plateau. Geology 38, 543-546.

ACC

EPTE

D M

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

34

Figure captions

Fig. 1: A. Location map of the Kisenge deposit (green star), in western Katanga (Democratic

Republic of the Congo), showing all major morphotectonic units mentioned in the paper and

used in Figure 3. B. Geological map of the study area (after Lepersonne, 1974); C.

Topographic map of the study area derived from Shuttle Radar Topographic Mission (SRTM)

digital elevation model (DEM).

Fig. 2: Schematic log of the Kis-1 profile, and of some other boreholes crossing the Kisenge

deposit (after Doyen, 1974); bedding is shown in grey shaded surfaces on the basis of dip

measurements. The dotted line (A) emphasizes the probable continuity between the bedding

and the position of the unweathered carbonate ore and garnetite level in the Kis-1 core. The

material studied in this paper comes from the Kis-1 core only.

Fig. 3: Overview of uplift and other major geodynamic events in the Congo Basin, Katanga,

the East African Rift System (EARS) and East African Plateau (EAP) during the Cenozoic,

with indications of supergene mineral deposits formed during this 45 My interval. Horizontal

dashed grey lines suggest possible correlations between events occurring in different

morphotectonic areas. The Mio-Pliocene interval in Katanga is further detailed in Figure 9.

Fig. 4: Backscattered electron images and SEM-EDS maps for selected samples. (A) RGM

13933, and (B) Kis-1/62 showing chemically relatively homogeneous zones; (C) Kis-1/36

displaying cryptomelane laminae with a minor Al-rich lithiophorite phase, confirmed by

XRD analysis.

Fig. 5: Pictures of selected studied samples: A to D are surface samples, E and F are taken

from the Kis-1 core. Inserts show the analysed and dated fragment (on a millimetre-sized

ACC

EPTE

D M

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

35

grid). Sample identification: A = RGM 1769; B = RGM 13200; C = RGM 10727; D =

RGM 14296; E = Kis-1/30; F = Kis-1/53 (see Table 2 for description).

Fig. 6: 39

Ar-40

Ar age spectra of Mn oxide samples. Age spectra of samples of the Kis-1 core

(Kis-1/30 to Kis-1/62) are grouped together for clarity, with indications of sampling depth.

Apparent age error bars are at the 1 level; errors in the J-values are not included. Plateau

and pseudo-plateau ages (1 uncertainties) are given when applicable. Frequency diagram of

apparent ages (grey) and of pseudo-plateau and plateau ages (coloured) are shown bottom

right. Identified phases are each indicated using a specific colour throughout the figure (these

colours are also used in Figures 8 and 9): >19.2 Ma (red), c. 15.7 Ma (orange), c. 14.2 Ma

(yellow), c. 13.6 Ma (purple), c. 10.5 Ma (blue), c. 3.6 Ma (dark green), c. 2.6 Ma (light

green).

Fig. 7: Processing of results for sample RGM 10727c with three identified phases α, β and γ.

A. Degassing diagram, presenting (xAr/ΔT°)/(

xAr/ΔT°)Max vs. %

39ArK, with x=40 for

40Ar*

(grey) and x=36 for 36

Ar (light blue), ΔT° corresponds to laser power increment; B. Inverse

isochron (correlation) diagram, with 36

Ar/40

Ar vs. 39

Ar/40

Ar, ellipses without fill are excluded

from isochron regression, MSWD stands for Mean Squares of Weighted Deviates; C.

Conventional age spectrum (apparent ages vs. %39

ArK); D. Weighted age spectrum presenting

apparent ages vs. %(39

ArK/ΔT°)/(39

ArK/ΔT°)Max; E. Photograph of sample grain RGM 10727c

taken before fusion step, with banded structures visible at grain surface.

Fig. 8: Selection of some inverse isochron (correlation) diagrams, plotting 36

Ar/40

Ar

vs.39

Ar/40

Ar; ellipses without fill are excluded from isochron regressions. MSWD stands for

Mean Squares of Weighted Deviates.

Fig. 9: A. Shuttle Radar Topographic Mission (SRTM) digital elevation model (DEM) of the

study area, with a topographic profile (orange line) from Kisenge (West) to Kolwezi (East);

ACC

EPTE

D M

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

36

three plateaus are observed, at ~1100 m a.s.l., ~1200 m and >1550 m a.s.l (Copperbelt); B.

Sketch of the three major cryptomelane formation phases (at 10.5, 3.6 and 2.6 Ma) with their

respective extent within the Kis-1 core. Each phase is superimposed on the previous one(s).

Colours for the three Mn oxide formation phases are also used in Figures 6 and 8.

ACC

EPTE

D M

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

37

Figure 1

ACC

EPTE

D M

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

38

Figure 2

ACC

EPTE

D M

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

39

Figure 3

ACC

EPTE

D M

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

40

Figure 4

ACC

EPTE

D M

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

41

Figure 5

ACC

EPTE

D M

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

42

Figure 6

ACC

EPTE

D M

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

43

Figure 7

ACC

EPTE

D M

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

44

Figure 8

ACC

EPTE

D M

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

45

Figure 9

ACC

EPTE

D M

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

46

Table 1: Overview of age information on Cenozoic supergene manganese (Mn) and

vanadium (V) ores deposits in western, central and southern Africa

Area Type Country Deposit Geologic context Radiometric

age (Ma)

References

West

Africa

Mn Burkina

Faso

Tambao Mn-laterite on primary

Proterozoic carbonate ore

Ar-Ar age

clusters:

- 59-56 - 47-44 - 27-24

Colin et al.

(2005);

Beauvais et al.

(2008)

Central

Africa

Mn D.R.Congo Kisenge Mn-laterites on primary

Paleoproterozoic (?) Mn-

carbonate (rhodochrosite)

ore

Ar-Ar ages:

- >19 - 11-10.5 - 3.7-3.5 - 2.7-2.6

Decrée et al.

(2010); this

paper

Southern

Africa

Mn N. RSA Kalahari

Manganese

Field

Altered manganese ore on

primary Paleoproterozoic

braunite lutite

Ar-Ar ages:

- 42 (max.) - 26.7 - 10.1 - 5.2

Van Niekerk et

al. (1999);

Gutzmer et al.

(2012)

Mn N. RSA Ryedale Ferromanganese wad

infilling in sediments

trapped within Permian

karsts on Neoearchean

dolomite

Ar-Ar ages:

- ca. 18-16 Pack et al.

(2000)

V N.

Namibia

Otavi

Mountainland

Zn-Cu-Pb vanadate

infilling in Miocene (?)

karst depressions within

sulfide-bearing

Neoproterozoic Otavi

carbonate

(U-Th)/He

ages:

- 47-1.5 (full

range)

- 33-24 (main

phase)

Boni et al.

(2007)

ACC

EPTE

D M

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

47

Table 2 : description of the analysed samples; depth along the core is given for the 5 core

samples. The „analysed feature‟ column is a description of the analysed and dated millimetric

fragments, while the „context‟ column describes the hand-size specimen whence the analysed

sample was taken. The „mould‟ character of spessartine is confirmed by the absence of this

mineral in XRD patterns.

Sample id Analysed feature Context

core

Kis-1/30

(depth -46m)

irregular cryptomelane veins or layers (up

to 2 cm thick) with metallic lustre

dark porous Fe/Mn-oxide-rich matrix with spessartine crystal

moulds; the cryptomelane veins are crossed by generation of

thinner (mm) Mn oxide veins

Kis-1/36

(depth -48m)

coalesced round cryptomelane aggregates

or nodules forming a cm-thick vein or

crust

mm-thick laminar coating at the interface with the dark grey to

brownish porous substrate with abundant spessartine crystal

moulds; crossed by mm-thick Mn oxide veins

Kis-1/53

(depth -56

m)

porous cryptomelane mass with numerous

moulds of spessartine crystals, but with

some massive mould-free zones

coatings or mm-thick veins; large irregular pores (up to 3 cm)

Kis-1/58

(depth -58

m)

coalesced round cryptomelane aggregates

(up to 5 mm diameter), with partially

concentric layering and metallic lustre

numerous inter-aggregate pores (up to 1 cm diameter)

Kis-1/62

(depth -64

m)

coalesced round cryptomelane aggregates

(up to 5 mm diameter), with partially

concentric layering and metallic lustre

dark dull cryptomelane matrix; sporadic large pores (up to 3

cm long)

surface

RGM 1767c cryptomelane crust (up to 1 cm thick)

with weakly expressed lamination and

metallic lustre

black spessartine-rich stratified substrate; some accessory

lithiophorite; the crust crosses the stratification at high angle

range (75° to 90°);

RGM 1769 heterogeneous cryptomelane mass with

contrasting parts with metallic and dull

lustre, commonly finely laminated;

probable spessartine moulds and larger

pores (up to a few mm)

botryoidal/mammillary surface, covered by layer (1 cm thick)

of prismatic crystals with a highly metallic lustre

RGM 10724 botryoidal/mammillary cryptomelane

crust (up to 3 cm thick) with weakly

expressed lamination

dark dense spessartine crystal moulds-rich black substrate

RGM 10727c complex botryoidal/mammillary

cryptomelane crust (up to 1 cm thick)

with clearly expressed laminated fabric

dark dense non-porous substrate

RGM 10739 cryptomelane porous crust (up to 4 cm

thick), with overall

botryoidal/mammillary aspect, consisting

of at least three superimposed palisadic

layers of wide prismatic units with inter-

aggregate porosity

none

RGM 13200 dense laminated cryptomelane crust (up base of the crust is irregular and contains fragments of the

ACC

EPTE

D M

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

48

to 6 cm thick) with metallic lustre, with

complex botryoidal/mammillary surface

underlying silty/clayey substrate

RGM 13201 cryptomelane crust (up to 3.5 cm thick),

weakly to clearly laminated, covered by

layer of parallel crystals (perpendicular to

surface)

none

RGM 13933 cryptomelane crust (0.5 cm thick), with

weakly expressed lamination, composed

of parallel crystals (perpendicular to

surface)

none

RGM 14296 cryptomelane vein (up to 1 cm thick),

composed of weakly to strongly

coalesced round aggregates, lined by

symmetric thin (~0.1 cm) dense Mn oxide

layers

quartzite matrix at both sides of the vein

RGM 17006 complex botryoidal/mammillary

cryptomelane crust (up to 1 cm thick)

with prominent lamination

porous spessartine crystal moulds-rich substrate; visible

intercalation of whitish silicates (confirmed with SEM-EDS)

in outer part of the crust

ACC

EPTE

D M

ANU

SCR

IPT

ACCEPTED MANUSCRIPT

49

Research highlights

- First extensive Ar-Ar study of Mn ores in Central Africa

- Main age clusters around 10.5, 3.6 and 2.6 My for supergene ore development

- Katanga as a key area for Neogene evolution of Central Africa